Australian government-funded trials still shut out many adults with intellectual disability
A retrospective analysis of the Australian New Zealand Clinical Trials Registry shows how often trial design can narrow access long before recruitment begins. Across 685 Australian government-funded interventional studies registered between 2022 and 2024, intellectual disability appeared in exclusion patterns both explicitly and indirectly, with consent capacity, diagnosis-based restrictions, and language requirements doing much of the work.
The most striking feature was not a single exclusion rule but the cumulative effect of many small ones. More than half of the included trials, 53%, contained no direct or indirect exclusion codes at all, yet 34.2% included at least one direct exclusion criterion and 25% included indirect exclusions. Those proportions matter because registry language often determines who is invited into research and who never gets the chance to be assessed as an individual participant.
Consent capacity drove the largest share of direct exclusions
Direct exclusion criteria were most commonly tied to consent capacity. In 17.2% of trials, eligibility language required a certain level of capacity to consent, a threshold that can become a blunt instrument when applied broadly rather than case by case. Diagnostic-label exclusions appeared in 13.1% of trials, while 4.8% explicitly excluded adults with intellectual disability.
In practical terms, the issue is not limited to one wording choice in a registry field. Capacity is often treated as though it were synonymous with diagnosis, yet adults with intellectual disability do not form a uniform group. Some can consent independently, some can consent with accessible information or support, and some may need substituted processes that still protect autonomy. When eligibility criteria rely on untested assumptions about cognition or understanding, they risk converting an ethical safeguard into a barrier to participation.
One sentence in the registry can determine whether a person is considered a candidate for research or excluded before conversation even begins.
Indirect exclusions were common and often hidden in plain sight
Indirect exclusion criteria appeared in a quarter of trials. English-language requirements were the most frequent, present in 12.3% of studies, followed by staff-discretion codes in 9.2%. These are not trivial administrative details. Language thresholds and discretionary screening can remove participants who may otherwise meet the scientific and clinical requirements of a study, especially when support strategies are not built into the protocol from the outset.
Indirect exclusion is particularly concerning because it can function as a proxy for disability without naming it. A trial may never mention intellectual disability directly and still exclude the people most likely to need adapted communication, additional time, or structured decision support. From a research ethics standpoint, that creates a gap between formal inclusion language and real-world access. The registry analysis captured that gap by separating direct exclusions from indirect mechanisms that can filter participants out at multiple points in the pathway.
Cancer and drug trials were the most restrictive, while larger studies showed a split pattern
Exclusion patterns were not uniform across trial types. Cancer and drug trials had the highest overall exclusion rates, suggesting that therapeutic area and intervention complexity influence how rigid eligibility criteria become. Larger trials enrolling 400 or more participants were least likely to contain any exclusions overall, yet they were more likely than small trials to explicitly exclude intellectual disability.
This pattern is clinically important because large trials often shape practice guidelines and policy decisions. If those studies systematically restrict enrollment through diagnosis-based or capacity-based criteria, the evidence base can drift away from the populations most likely to encounter those treatments in routine care. In cancer, where symptom burden, treatment toxicity, and communication needs already complicate participation, narrow eligibility can magnify existing inequities. Drug trials may also lean heavily on standardized consent language and protocolized endpoints, which can make investigators more likely to treat communication support as a logistical problem rather than a design requirement.
The resulting evidence base may look clean on paper and still fail to reflect the patients clinicians see every day.
Registry language reflects ethics, but it also exposes where ethics are not yet built in
The study was designed as a retrospective analysis of ANZCTR records for interventional trials funded by an Australian government source and registered between 1 January 2022 and 31 December 2024. Trial records were screened carefully, with phase 0, 1, 1/2 and 2 studies excluded, along with trials whose participant age thresholds fell outside the adult range examined here. Of 857 screened records, 685 met inclusion criteria.
Eligibility criteria were extracted from free-text registry fields and coded with an adapted framework that distinguished direct exclusions from indirect ones. That distinction is useful because registry language often compresses complex consent, communication and diagnostic considerations into a few short phrases. By translating those phrases into structured categories, the analysis made visible how often adults with intellectual disability may be filtered out before a trial can even test whether support measures would have made participation possible.
There is a clear ethical tension here. Trials that exclude participants on the basis of assumed consent or cognitive capacity, without individualized assessment, sit uneasily with guidance favoring supported decision-making and tailored accommodation. In that sense, the registry is not just a database. It is a record of what investigators believe participation should require.
Eligibility design could become more inclusive without weakening scientific rigor
The findings point toward practical changes in trial design rather than abstract concern. Eligibility criteria should justify diagnostic and capacity-based exclusions explicitly, rather than using them as default filters. Protocols can also incorporate validated methods to support participation of adults with intellectual disability, including accessible consent materials and individualized assessment pathways. If a study genuinely requires a particular level of cognitive function or communication for safety or endpoint validity, that requirement should be stated narrowly and tied to the scientific purpose of the trial.
This matters because broad exclusions are not neutral. They shape who benefits from publicly funded research, who contributes to the evidence base, and who remains invisible in the data. A registry filled with vague references to capacity or language can look methodologically tidy while quietly narrowing the population to those easiest to enroll.
For Australian government-funded research, the remaining gap is not whether inclusion is possible in principle. The more pressing question is whether trial teams will build it into eligibility design from the start, rather than treating adults with intellectual disability as exceptions to be managed at the margin.